I here advocate an approach to linguistic fieldwork in which text collection and elicitation are interwoven in a finely tuned, constantly modulated way. By text collection I am referring to the practice of compiling and analyzing naturally occurring speech and narratives in the language under study; by elicitation I mean either the use in language analysis of native-speaker intuitions or translations of decontextualized utterances from a contact language to the language being studied. Both practices are well motivated: text collections are reservoirs of cultural and linguistic information, and elicited forms provide crucial evidence necessary for the formulation of grammatical generalizations. In my own fieldwork experience with Meithei, a Tibeto-Burman language spoken in Northeast India in Manipur state, I have found that linguistic generalizations that result exclusively from elicitation tend to be unreliable. Likewise, language description based solely on textual data results in patchy and incomplete descriptions. I conclude that reliable and usable field data can only be collected when both text collection and elicitation are used.

Using text collection in conjunction with elicitation

My investigation of Meithei began, predictably, with the elicitation of wordlists and minimal pairs. I used such elicitation to build a rudimentary understanding of the phonology of Meithei. I then attempted the elicitation of paradigms of verb conjugations and noun declensions in order to discover the basic inflectional morphology. I studied case marking and word order through the elicitation of simple sentences with verbs that could be expected to require one, two, or three arguments.